Analysis of Regime Switching Behaviour of Indian Stock Markets

I consider the problem of detecting and predicting regime switching behaviour in the context of Indian Stock Market data. First I discuss detection of volatility change points using the LRT and the Binary Segmentation procedure of Vostrikova (1981). The detected volatility changes are correlated with important national and international events during that time. Then I analyse regime switching behaviour using Hidden Markov Models. I use a Bayesian route and use the Sampling Importance Resampling methodology to obtain the approximate posterior distributions